Abstract
This chapter is devoted to the application of virtual measures in methods of standardless diagnostics of composites, which allows to simulate various types of defects or damage levels of a product without the use of physical standards, which reduces the time, technical and economic costs associated with the manufacture of reference samples. A method for constructing virtual measures (standards) that can be used to train and configure diagnostic systems, assess the reliability of diagnostic methods is considered. The method allows to synthesize a model of the information signal of the flaw detector, which takes into account the deterministic and random components of the characteristics of real signals. The use of neural networks in standardless diagnostic systems allows not only to recognize the information signals received during control, but also to store information about the patterns and the relationship of the characteristics of the information signal and the state of the control object, provides the ability to correctly classify information signals corresponding to possible defects that were not encountered during training network.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
